Why OBAMA will lose PA (Pflouffe gives away faulty turnout assumptions)

Fox News | 10/24/08 | Thane Banquo

[Thane_Banquo]

Earlier this evening FoxNews ran a piece quoting David Pflouffe (read "poof") saying that the Obama camp couldn't figure out why McCain was campaigning in PA. Pfloufe said that in order for McCain to win PA, he would have to win 20% of Democrats, 95% of Republicans, and 60% of Independents.

So I calculated, based on 2004 PA turnout of 41% Democrats, 39% Republicans, and 20% Independents, what kind of edge that would give McCain.

If McCain managed to pull off such a feat, he would win 57%-43%, a 14 pt margin! If McCain wins PA by 14 pts, he'd probably win every other state but Illinois!

So then I decided to find out the DEM-GOP spread necessary to come up with Pflouffe's scenario. In other words, I wanted to see what kind of DEM turnout Pfloufe was expecting in order to have McCain just barely edge out Obama by winning 20% of D's, 95% of R's, and 60% of I's

In order for McCain to actually need that kind of result, Pfloufe must be expecting Dem turnout to be roughly 20 pts higher than GOP turnout in Pennsylvania!

If they actually think that's going to happen on election day, the Obama camp is going to be very disappointed.

The point? Now we see who is feeding these faulty Dem turnout assumptions to the major pollsters

[Alter Kaker]

The polling samples are unbelievable. Nearly every second person in the country is an Obamabot.

You're suprised? That happens every election. In 2004, 48% voted for Kerry. In 2000, 49% voted for Gore.

[Norman Bates]

Wow - that’s close. Just a little more.

[comebacknewt]

Some polls are deliberately misleading IMO.

The majority are sincere in their efforts, but differ widely because of their assumptions about the true makeup of the electorate and because of their turnout models.

If RATs do outnumber us by 6%+ on election day, and if the Obama turnout machine is as great as advertised, then it will be a landslide victory for the RATs. If either one of those are not true, we are looking at a much closer race.

If both prove not to be true, we could be looking at President McCain. He is a decided long-shot, but every once in a while they long-shots do come in.

[1035rep]

It’s the fake ACORN registrations.

[Yaelle]

I am usually the only McCain car in the Whole Foods (in SoCal) parking lot. Yet today there were two of us. Hope and change!

[kabar]

Kerry received 10,000 more votes than Bush from Erie county in 2004.

[Thickman]

The early exit polls were wrong. Early exit polls are not the same thing as telephone polls. Pretending that they are is nuts. The methodology is completely different.

Calling Captain Obvious . . . of course they are different. Exit polls are supposed to be more accurate because instead of predicting if someone is going to vote, the poller knows the subject has just cast a ballot. Get the point? Maybe not.

[kabar]

Turnout in 2004 was much greater than the 2006 midterms as is always the case in Presidential years. Turnout wins elections.

[Thickman]

So you think that at some point since 2004 every single polling organization in the United States decided to change its polling technique? Are they supposed to have done this in a cave outside Zurich or something?

I don't think it, I know it because I read the internals. You might try that sometime. Be careful - failure to examine below the surface of things is a troll trait.

[cookcounty]

I wish someone would do the research on the polls in the primaries. It is my strong impression that Obama usually polled much higher than the actual result, but don’t have the data at my fingertips to check.

[kabar]

The Dems have done a good job of registering new voters. The question is whether this will translate into more Dem votes or not. I assume the pollsters changed the weighting to reflect the new registration numbers. The unknown factor is how much of this is due to ACORN and Reps/Independents shifting during the highly competitive Dem primaries.

[kabar]

Just the opposite. Obama usually polled at or below the number. Rarely did he exceed the poll forecasts.

[Alter Kaker]

Exit polls are supposed to be more accurate because instead of predicting if someone is going to vote, the poller knows the subject has just cast a ballot.

Supposed to be? Supposed to be by whom? Thickman?

In fact, the exit polls in '04 were fraught with error for reasons that are well understood by political scientists.

The vast majority of poll checkers were college students and they had way too much leeway in picking interviewees. As a result, the exit polled universe in '04 skewed heavily towards the kinds of people who would be eager to talk to college students -- a population considerably more Democratic than the overall vote universe.

There have been numerous papers written on this.

That error is unique to face-to-face exit polling, and has nothing to do with telephone polling. In fact, the telephone polling in 2004 was, on the whole, extremely accurate.

[Alter Kaker]

I don't think it, I know it because I read the internals.

So you believe that every single polling organization in the country is in the tank for Obama? Every one? Not one refused to sell out their methodology? I read polls' internals too. But I also try to understand them, something I don't think you do.

Be careful - failure to examine below the surface of things is a troll trait.

Careful who you call a troll. I've been on FR for longer than you have, pal.

[Thickman]

So you believe that every single polling organization in the country is in the tank for Obama? Every one? Not one refused to sell out their methodology? I read polls' internals too. But I also try to understand them, something I don't think you do.

Typical superficial comeback. Did I say every one? No! In fact, there are three respected polls that currently have the race at 1, 2 and 3 points. I'm referring only to the polls you are citing (which reinforce your argument but are the ones that ARE in the tank as evidenced by their oversampling of dummies).

Careful who you call a troll. I've been on FR for longer than you have, pal. Reading is fundamental. I did not call you a troll. I do stand by the assertion that your observations are troll-like. Pal.

[Chet 99]

You’re over analyzing the situation.

If McCain wins PA, he’s going to win VA, NC, IN, MO, etc.

If McCain loses IN or NC, it will be a nationwide route. There is talk of 50 separate contests, but the states generally move in tandem with the national popular vote. There are exceptions of course, but I don’t think you’ll see that this year with PA.

[Thickman]

Thanks for supporting my point. Now that we agree on the fact that exit polling was wrong, can we address the issue that you keep avoiding? Namely, that the methodology of the polls you cling to (as opposed to others that don’t show Barry way ahead) might be flawed by oversampling dummies, African-American, urban dwellers and youngsters? That was a rhetorical question because after several vacuous posts by you I don’t expect an answer.

[Bush Revolution]

Yeah, my head hurts...I have to go read that again.

[Bush Revolution]

More stories like this please, they make my day.

[Alter Kaker]

The difference is that in '04, the exit polling consisted of one single consortium effort, using a single methodology.

If you're criticizing 2008's telephone polls, then you're going to be criticizing different methodologies by several dozen different independent organizations. For the most part, those methodologies have been accurate in the past -- and were in 2004. Now there are instances where the methodologies are clearly unusual -- Zogby's Internet Polling being the obvious one -- but for the most part the sample methods have been consistent election to election. And the results too.

That was a rhetorical question because after several vacuous posts by you I don’t expect an answer.

I don't know what your problem is but you have one hell of an attitude.